Aluminum as rocket fuel?

[Raukk]

Aluminum as rocket fuel?

This discussion is purely theory, and for fun.

Please point out any errors: math/logic/etc.

Note: numbers are from Wikipedia unless otherwise stated.

At first glance Aluminum seems like a terrible Rocket fuel, but after I ran some numbers it appears that it could compete with LH2. I’m aware of other projects using Aluminum, like AlIce: http://en.wikipedia.org/wiki/ALICE_(propellant) But I just want to talk about theoretical ideas with this thread.

Initially the Energy Density of LH2 of 141.86 MJ/Kg versus Aluminums 31 MJ/Kg seems like an obvious winner except that for a rocket the Oxidizer must also be carried on board and should also be included into that equation. For every 2 atoms of H we need one atom of O; 2*H + O = H2O : 2*1 + 16 = ~18 molecular weight, so for each Kg of LH2 we need 8Kg of LOX. 141.86 / 9 = 15.762 MJ/Kg when the LOX is included in the weight.

Interestingly Aluminum being reduced to AL2O3 smaller Oxygen requirement; 2*AL + 3O = Al2O3 : 2*27 + 3*16 = ~102 molecular weight, so for each Kg of aluminum only .889 Kg of LOX is needed. Which means 31/1.88 = 16.41MJ/kg is the energy density per Kg of fuel + LOX, which is higher than LH2.

Looking at these numbers I assume that I made a mistake somewhere in the math or science somewhere.

The other Thing I find interesting is that Aluminum is almost 40 times denser than LH2, so that when you consider an equal amount of energy, say 10,000 MJ (about 1 cubic Meter of LH2). LH2 would take 70.85 Kg and 566.8Kg of LOX, LOX is 1141 Kg/cubic meters, so 566/1141 = ~.5. The total LH2+LOX fuel has a volume of 1.5 cubic meters. Aluminum needs 323 Kg to create that amount of energy, at 2700 Kg/cubic meter that is a volume of 0.12 cubic meters; it also requires 287Kg of LOX, giving an additional .251 cubic meters for a total of .371 cubic meters. Which gives us a ratio of 1.5 to .371 gives us a total of 4 times as much volume in fuel for LH2 as for Aluminum.

I am sure there are huge technical hurdles to making Aluminum a Rocket Fuel, but if it has the same energy density as LH2 but ¼ the volume I would expect it to be under more significant development.

My current conclusion is that I have missed something or done some math incorrectly or else we would see more research into making rockets powered by Aluminum.

What are your thoughts on this?

If my logic and Math is correct, what do you think are the technical hurdles that would make Aluminum non-viable?

[Kryten]

The much higher weight of aluminium oxide relative to water means you end with lower exhaust velocity, and thus ISP, despite the higher theoretical density. To be more specific, you get about 2,800m/s (285s ISP), compared to 4,600m/s (465s) for hydrolox.

[Guest]

While it's a shoddy propellant for launching from Earth, a theoretical Moon base would find it great - aluminium is literally laying around the Moon, meaning it'd be very cheap to extract and use in suborbital "hoppers", communication with orbit and Earth return. Low efficiency wouldn't be an issue due to the fact you don't need a lot of dV to get off the Moon.

[sal_vager]

Hasn't this been tried with SRB's?

[Raukk]

The much higher weight of aluminium oxide relative to water means you end with lower exhaust velocity, and thus ISP, despite the higher theoretical density. To be more specific, you get about 2,800m/s (285s ISP), compared to 4,600m/s (465s) for hydrolox.

Interesting, might I inquire where you got both those numbers from?

[Kryten]

Pulled from here. Not checked it, but it sounds right. If you wanted to do it from scratch, you'd have to calculate via working out the kinetic energy per product molecule.

[Tex_NL]

Interesting, might I inquire where you got both those numbers from?

http://www.permanent.com/lunar-geology-minerals.html talks about Al₂O₃ concentrations between 10.32 and 27.18%

[Tex_NL]

Interesting, might I inquire where you got both those numbers from?

http://www.permanent.com/lunar-geology-minerals.html talks about Al₂O₃ concentrations between 10.32 and 27.18%

So yes, you're pretty much 'swimming' in rocket fuel as long as you got the tech to separate the aluminium and oxygen.

[lajoswinkler]

While it's a shoddy propellant for launching from Earth, a theoretical Moon base would find it great - aluminium is literally laying around the Moon, meaning it'd be very cheap to extract and use in suborbital "hoppers", communication with orbit and Earth return. Low efficiency wouldn't be an issue due to the fact you don't need a lot of dV to get off the Moon.

Aluminium is not laying around the moon. Aluminium compounds are. They are extremely difficult to process to isolate the metal out. Aluminium producing industry is one of the most intensive consumers of electricity.

[Raukk]

Pulled from here. Not checked it, but it sounds right. If you wanted to do it from scratch, you'd have to calculate via working out the kinetic energy per product molecule.

Thank you, that's quite interesting. I'll have to look around that site more, I really wish they would show their math/source.

I'm going from the simplified idea that if you pump 15 Mega-Joules into 1 Kg of material regardless of it's density it would still exit at the same speed E= 1/2 M V^2. I'm probably not taking into account the Ideal Gas law PV=NRT, where 1KG of H2O would have more moles than 1Kg Al2O3 and then the higher pressure would make a higher exit velocity.

[Kryten]

Hasn't this been tried with SRB's?

That's aluminium with ammonium perchlorate, it's rather more complex.

3NH₄ClO₄+3Al--->Al₂O₃+ AlCl₃+3NO+6H₂O

[Raukk]

http://www.permanent.com/lunar-geology-minerals.html talks about Al₂O₃ concentrations between 10.32 and 27.18%

I'm not sure why this is a quote of my comment, I was asking for the source of his ISP numbers, not how much of it exists on the moon.

I'd point out that we are swimming in Aluminum too. "Aluminium is the third most abundant element (after oxygen and silicon), and the most abundant metal in the Earth's crust."

[Kryten]

Thank you, that's quite interesting. I'll have to look around that site more, I really wish they would show their math/source.

I'm going from the simplified idea that if you pump 15 Mega-Joules into 1 Kg of material regardless of it's density it would still exit at the same speed E= 1/2 M V^2. I'm probably not taking into account the Ideal Gas law PV=NRT, where 1KG of H2O would have more moles than 1Kg Al2O3 and then the higher pressure would make a higher exit velocity.

If you are just throwing a 1KG brick of stuff out the back you'd get the same performance for the same energy regardless of the exact kind of stuff, but in practice you're throwing out millions of smaller bricks. The two processes aren't equivalent because one is a continuous process while the other is a single event; you have to use the rocket equation, not simple kinetic energy. The reason basic kinetic energy calculations don't work is not all of that kinetic energy will enter the ship; the fuel from the last part of the burn will carry with it kinetic energy that was given to it at the start of the burn.

[Raukk]

Yep, that's what I get for over simplification.

[peadar1987]

Aluminium is not laying around the moon. Aluminium compounds are. They are extremely difficult to process to isolate the metal out. Aluminium producing industry is one of the most intensive consumers of electricity.

I would imagine that if we're in the market for ISRU and suborbital transportation on the lunar surface we'll probably be able to produce lots of energy relatively easily. Light, efficient solar panels and the like.

[lajoswinkler]

I would imagine that if we're in the market for ISRU and suborbital transportation on the lunar surface we'll probably be able to produce lots of energy relatively easily. Light, efficient solar panels and the like.

Energy requirements for production of aluminium is insane. We're talking about hundreds of kiloamperes of current at few volts of operation. That's a lot of parallel solar panels. I mean a lot.

What about heat management? Hall–Héroult process goes at some 1000 °C and deals with evaporating fused coke anode. The heat involved is enormous and dealing with it in vacuum is a PITA. It's a chemical engineer's nightmare.

[KSK]

Energy requirements for production of aluminium is insane. We're talking about hundreds of kiloamperes of current at few volts of operation. That's a lot of parallel solar panels. I mean a lot.

What about heat management? Hall–Héroult process goes at some 1000 °C and deals with evaporating fused coke anode. The heat involved is enormous and dealing with it in vacuum is a PITA. It's a chemical engineer's nightmare.

Interesting paper here on regolith refining. It proposes fluorine processing (transported as potassium fluoride) of the regolith, distillation to separate out the metal fluorides and then reduction with potassium to regenerate the potassium fluoride and release your metals (iron, titanium, aluminium). The method was primarily designed for silicon processing for solar cell manufacture.

[CluelessModeler]

That's aluminium with ammonium perchlorate, it's rather more complex.

3NH₄ClO₄+3Al--->Al₂O₃+ AlCl₃+3NO+6H₂O

True!

Actually, the Space Shuttle SRB's used Aluminum as their "primary fuel."

It's complemented by Ammonium Perchlorate, which serves as an oxidizer. Along with Iron Oxide, a sort of a catalyst, because it's easier to ignite. Along with a couple of polymers and an epoxy resin that cures the propellant in place.

This is all pulled from Wikipedia, but, it seems legit.

[lajoswinkler]

Interesting paper here on regolith refining. It proposes fluorine processing (transported as potassium fluoride) of the regolith, distillation to separate out the metal fluorides and then reduction with potassium to regenerate the potassium fluoride and release your metals (iron, titanium, aluminium). The method was primarily designed for silicon processing for solar cell manufacture.

Thanks, I'll check that out.

[Raukk]

That's aluminium with ammonium perchlorate, it's rather more complex.

3NH₄ClO₄+3Al--->Al₂O₃+ AlCl₃+3NO+6H₂O

I assume the reason they use ammonium perchlorate rather than a more Oxygen dense Oxidizer is for the multitude of molecules that this reaction creates? Wouldn't that mean a higher pressure at a lower temperature, I assume Al + O2 would just melt the rocket?

When compared to my earlier calculations this reaction has a much worse MJ/Kg so it must have some other benefit.

If anyone knows the official reason that they use that combination then I'd be curious to hear it.

[Kryten]

Lower average molecular weight for products leads to higher exhaust velocity. Also, it's all-solid, so there's no expensive plumbing to worry about.

Edited October 22, 2014 by Kryten

[KerikBalm]

Hmmm, at the beginning, I was thinking that, but I think the OP has a point.

If we were talking NTR here, then we'd have a maximum temperature, which translates to a KE for each gas molecule.

To get the highest exhaust velocity, witha fixed KE, you want the lowest mass per molecule.

But this is not an NTR. Like an NTR, for a given molecular weight, if you want a higher exhaust velocity/ISP, you need a higher temperature.

Thus hypothetically, you can get a higher ISP from a higher molecular weight gas if that gas is a lot hotter.

Ie, run one NTR with a core temperature of 500K, expelling H2O gas, and another NTR with a core temperature of 3,200k, expelling N2 gas.

Which gas gets the higher ISP?

With a chemical reaction such as this, we have to consider how hot the reaction burns, which is related to how much energy the reaction releases.

If the products have 2x the molecular weight, but 4x the energy, they should have the same velocity and the reaction should give you the same ISP.

[Raukk]

@KerikBalm: That is what I was thinking by the end of it, but I'm guessing that the physical constraints of the combustion chamber to deal with the 4x temperature would be to great to overcome (or would make LH2 more viable).

Edit: If my initial (and probably wrong) calculation is right then the Al2O3 would need to be at 5.6 times the temperature of the H2O leaving the engine. That means that you are dealing with 10,000-20,000 degrees C for the engine, I don't think we have any material that can deal with that.

As was mentioned with the SRBs, it's a much different reaction that carries a lot more mass per unit energy but gives a lot more (and lighter) resultant compounds which would mean a lower operating temperature.

After this very informative thread I have come to the conclusion that in theory an aluminum rocket would be as effective or more than LH2, but in reality constructing such a rocket is not possible/economical, especially when LH2/LOX is so common and well understood.

Edited October 22, 2014 by Raukk

[KSK]

I don't know the official reason but there must be a good one, simply because perchlorates are Not Nice and aluminium chloride isn't a great thing to be spewing into the air in great quantities either. I suspect (with absolutely nothing to back this up) that it's used precisely because it is a pretty oxygen dense oxidiser. I'm sure someone will have a neat counterexample but off the top of my head I'm struggling to think of a better one that is also a solid.

[KSK]

Thanks, I'll check that out.

You're welcome! No idea how legit it is, but the authors do at least seem to have considered some of the problems - and I liked the opening statement in section 2.1 that: "Extraction of purified feedstocks from the lunar regolith cannot be done using the industrial production processes currently in use on Earth."